CA2473243A1

CA2473243A1 - High surface area magnesia curing agent

Info

Publication number: CA2473243A1
Application number: CA002473243A
Authority: CA
Inventors: Jimmy Pingao Chen
Original assignee: Individual
Current assignee: Hexion Inc
Priority date: 2002-01-17
Filing date: 2003-01-16
Publication date: 2003-07-31
Also published as: CN1617778A; WO2003061871A3; US6710101B2; KR20040083076A; WO2003061871A2; MXPA04006918A; AU2003225519A1; US20030153648A1; TW200304471A; BR0307174A

Abstract

Binder compositions of the present invention are useful in refractory and foundry applications. The compositions include an aggregate, a liquid resole resin, and a magnesium hardening agent. The compositions are workable and exhibit high compressive strengths in a short period of time. Additionally, the compositions have a low free phenol content and require a reduced resin content as compared to currently known compositions. The compositions reduce the cost to produce useful binders by eliminating the requirement for ester caning in addition to requiring reduced amounts of resin to obtain adequate compressive strength.

Description

HIGH SURFACE AREA MAGNESIA CURING AGENT
FIELD OF THE INVENTION
The invention relates to a room temperature curing binder for refractory or foundry applications. The binder incorporates lightburned magnesium oxide particles.
BACKGROUND OF THE INVENTION
The control of the ambient temperature curing of binder compositions is useful in the foundry and refractory arts. In foundry and refractory applications, the binder is mixed with sand or other refractory material and various shapes are formed with the help of pattern equipment. Room temperature or ambient temperature hardening systems, used in foundry and refractory applications depend on their ability to uniformly coat sand or other refractory grains which then cure into strong, rigid shapes at ambient conditions.
One room temperature cure process, lazown as ester cure, uses a highly allcaline phenolic resole resin with a pH greater than 11 and an alkali to phenol molar ratio of 0.2/1 to 1.2/l. Generally, the binder components are mixed into sand in either a batch or a continuous process and the mixed sand is discharged into a pattern. The sand begins to cwe or harden immediately and it is essential to sufficiently pack the sand to achieve optimum bonding strength. If the sand cannot be sufficiently paclced prior to curing there is a diminished bonding strength. The usefulness of a binder is related to the amount of time available to sufficiently work the binder into a pattern prior to curing of the binder.
Compositions for retarding the ambient temperature hardening of a phenolic resole resin alone or with an aggregate when such resin is contacted with a nitroallcane and a hardening agent such as lightburned magnesiLUn oxide have been used. The pH
of the phenolic resole resin used in that application varied over a broad range fiom about 4.5 to 13. However, hardening takes place at a pH above 7, i.e., in the allcaline range such as that above 7.5. When the hardener is lightburned magnesium oxide alone or together with an ester functional hardening agent, the pH of the resin can be below 7 such as between a pH of 4.5 and 7, but a sufficient amount of the lightburned magnesium oxide needs to be present to both neutralize the acidity and to provide sufficient magnesium oxide for the crosslil~lcing and hardening of the resin.
Ambient temperature hardening of compositions containing magnesia aggregate and a curable, liquid phenolic resin, either alone or together with an ester function hardening agent, has been accelerated by the use of additives such as those which supply: acetate; adipate; 1,2,4-benzenetricarboxylate (trimellitate);
fonnate;
glycolate; lactate; nitrate; succinate; sulfamate; phenolsulfonate; or toluenesulfonate alllo115 to the CO111pOSltlOn Or COlnpOUlldS WhlCh Supply acetylacetone (2,4-pentanedione); 2-nitrophenol; 4-nitrophenol; or salicylaldehyde to the composition.
This curing method has been used with resole resins having a molar ratio of phenol-to-formaldehyde of between about 1:0.9 and 1:3. However, the phenol-to-formaldehyde ratios on the lower end of this range do not produce high strength binders in a short period of time. A novolac resin could also be used as a liquid solution if used alone as the phenolic resin or as a liquid or solid when used together with a resole solution.
Lightburned magnesilun oxide products having different surface areas can be obtained from various sources such as the Martin Marietta Magnesia Specialties Company, Baltimore, MD, under the designator of MAGCHEM Magnesium Oxide Products. Lightburned magnesium oxides with the higher surface areas are more active and provide shorter times for gelation and hardening. Reactivity and surface area of magnesium oxide (lnagllesia) differ greatly depending on the procedure used for manufacture of the magnesia. Thus, lightburned magnesia has a surface area of about 10 to 200 or snore square meters per gram. Hardburned magnesia and deadburned magnesia have a surface area of about one or less than one square meter per gram.
For use in refractory compositions, the magnesia grain has been crushed and sized in various fractions. Commonly used grain sizes of deadburned or hardburned grades of magnesia have been used for room temperature hardening, meaning the hardening of binder-aggregate compositions tools place at temperatures of about 60° F
to 90° F.
IW own binder-aggregate compositions produced by combining a curable resin binder, magnesia aggregate, and accelerator have additionally comprised a number of optional modifiers or additives including: non-reactive solvents; silanes;

hexamethylenetetraamine; clays; graphite; iron oxide; carbon pitch; silicon dioxide;
metal powders such as aluminum, magnesimn, and silicon; surfactants;
dispersants;
air detraining agents; and mixtlues thereof.
There is a need for a refractory mixture that clues at roo111 temperature that provides adequate work time and then cures quickly and has sufficient compressive strength. There is fwther a need for such a refractory composition that provides a lower free phenol content and can be produced at a lower cost due to a decrease in the required resin concentration.
SUMMARY OF THE INVENTION
The present invention provides refractory compositions that minimize resin content. The advantages related to decreased resin concentration in the refractory mixtLUes are two-fold. First, the product cost is reduced, and secondly, there is a reduction in emissions associated with the resin. The compositions are workable and exhibit high compressive strengths in a short period oftime. Additionally, the refractory mixtures of the present invention have a low free phenol content and require a reduced resin content as compared to compositions of the prior art.
The refractory mixtures provided in the invention fitrther reduce the cost to produce useful al-ticles by eliminating the requirement for ester curing in addition to requiring reduced a11101111t5 Of re5111 to obtain adequate compressive strength.
The present invention is directed t0 a C0111pOSltlOn 111C1Lldlllg a liquid resole having a mole ratio of phenol to formaldehyde ranging from about 1:2.0 to about 1:2.4; an aggregate; and a magnesium hardening agent.
DETAILED DESCRIPTION OF THE INVENTION
According to one embodiment of the present invention there is provided a room temperature curing composition using varying concentrations of lightburned magnesium oxide. The compositions of the present invention use a hquld resole having a specified mole ratio of phenol-to-formaldehyde. It has been discovered that the use of the liquid resole of the present invention in combination with the lightburned magnesium oxide described herein provides surprising and unexpected rates of hardening and compressive strength development as compared to prior art compositions using ester, phenolic resole and lightbluned magnesium oxide.

In another embodiment of the present invention there is provided a binder composition, including:
an aggregate;
a liquid resole having a mole ratio of phenol to formaldehyde ranging from about 1:2.0 to about 1:2.4, in an amount ranging fiom about 1% to about 20% by weight, based on the total weight of the aggregate;
lightburned magnesiwn oxide in an amolmt ranging from about 0.5%
to about 50% by weight, based on the total weight of the liquid resole.; and a solvent in an amowt ranging from 0% to about 25% by weight, based on the total weight of the aggregate.
The term "hardening agent" is used herein to denote a material which increases the rate of hardening of a phenolic resole resin, e.g., at room or ambient temperature (R.T.). Hardening is attained with increases in viscosity and gelation to form a solid that is firm to the touch and generally inflexible. An example of a lightburned magnesium hardening agent is lightburned magnesium oxide.
By the term "room temperatlue hardening" we mean the hardening of CO111pOS1t1011S Of thlS 111Ve11t1011 at te111pe1'ature Of about 60° F.
to 90° F., particularly about 65° F. to 80° F.
The magnesium hardening agents are magnesium hydroxide, lightburned magnesiLUn oxide, or other magnesium oxide which has the hardening activity for phenolic resole resins of lightburned magnesilun oxide such as that having a surface area of at least 10 square meters per grain (10 mz/g).
Reactivity and surface area of magnesium oxide ("magnesia") differ greatly depending on the procedure used for manufactLlre of the magnesia. Lightburned grades of magnesium oxide are calcined at temperatures ranging from about 1600° to 1800° F. I-Iardburned grades are calcined at temperatures ranging from about 2800°
to 3000° F. Deadbu reed or periclase grade of magnesium oxide is calcined at temperatures of over 4000° F. The lightburned grades are generally available in powder or granulated form while hardbu rued grades are available in kiln rim, milled, or screened sizes. Periclase is generally available as briquettes and as screened or milled fractions. There are large differences in surface areas for the various magnesias. Thus, lightbluned magnesia has a surface area of about 10 to 200 or more square meters per gram (m2/g). I-Iardburned magnesia has a surface area of about one square meter per gram, whereas deadburned magnesia has a surface area of less than one square meter per gram. Magnesia which is conventionally used as a refractory aggregate is the deadburned or periclase magnesia. Neither hardburned nor deadburned magnesia are effective hardening agents. It is the lightburned magnesia which is an effective hardening agent. Lightblirned magnesia products having different surface areas can be obtained from the Martin Marietta Magnesia Specialties Company, Baltimore, MD, under the designator of MAGCHEM Magnesilun Oxide PTOCILICtS. Illustratively, MAG-CHEM 30 has a surface area of about 25 square meters per gralll. MAGCHEM 50 has a surface area of about 65 square meters per gram whereas MAGCHEM 200D has a surface area of about 170 square meters per gram.
In the present invention, the amount of lightburned magnesia to be used is dependent on the surface area of lightburned magnesia employed. For example, comparatively less MAGCHEM 200D would be used than MAGCHEM 50, and less MAGCHEM 50 would be used than MAGCHEM 30.
One of the variables influencing viscosity increase, formation of gel and subsequent hardening of a phenolic resole resin is the surface area of the lightburned 111ag11eS111111 OXlde. Magnesium oxides having higher sLlrface areas are more active and provide shorter times for gelation and hardening. Thus, lightburned magnesilun oxide, having a surface area of less than about 25 square meters per gram, is slow acting and generally will not be used when it is desired to have the binder composition cure in a relatively short period of time at temperatures below about 120°
F. On the other hand, magnesia having a higher surface area, such as about 65 square meters per gram ~1112~g~ and above, will harden the same binder composition in a shorter period of time. For many applications, using magnesia having a surface area of about 25 to 65 square meters per gram is suitable. Hardburned magnesia reacts too slowly as a hardener to be of practical value, and deadburned magnesia is sufficiently inert so that it is used conventionally as a refractory with phenolic resin binders with little or no effect on room temperature hardening rates.
The quantity Of llghtbLlrlled 111ag11eSltull Oxlde Or lnaglleSltlln hydroxide which is used in 11118 111Ve11t1011 aS a hardelleT 1S all allloullt sufficient to increase the rate of gelation or hardelllllg of the phenolic resole resin.
Preferred phenolic resole resins used in this invention have less than about 2%
by weight of water soluble sodilun or potassium. A preferred molar ratio for use in this invention ranges from about 1 mole of the phenol for each 2.0 moles of the aldehyde to about 1 mole of phenol for each 2.4 moles of the aldehyde and particularly a range of phenol to aldehyde of about 1:2.1 to about 1:2.3. The phenolic resole resin will usually be used in solution.
The pH of the phenolic resole resin used in this invention will generally range from about 8 to about 9, a preferred range being from about 8.5 to about 9.
The liquid portion of the resin is water or water together with a non-reactive solvent. The resin can include a wunber of optional modifiers or additives such as silanes, hexamethylenetetramine, or urea. Solvents useful for the present invention in addition to water can be selected from alcohols of one or five carbon atoms, diacetone, alcohol, glycols of 2 to 6 carbon atones, mono- and dimethyl or butyl ethers of glycols, low molecular weight (200-600) polyethylene glycols and methyl ethers thereof, phenolics of 6 to 15 carbons, phenoxyethanol, lactones such as 'y-butyrolactone, y-valerolactone, and 8-valerolactone, aprotic solvents such as N,N-dimethylformamide, N,N-dimethylacetamide, 2-pyrrolidinone, N-methyl-2-pyrrolidinone, dimethyl sulfoxide, tetramethylene sulfone, hexamethylphosphoramide, tetramethyl urea, methyl ethyl lcetone, methyl isobutyl lcetone, cyclic ethers such as tetrahydrofuran and m-dioxolane, and the like, and mixtwes thereof. Under certain conditions of elevated alkalinity, esters and lactones such as y-butyrolactone may be reactive with the resole.
Preferred water contents for the resins used in this invention will vary from about 18% to about 24% by weight of the resin and can thus be referred to as aqueous SolL1t1011S.
Organofunctional silane adhesion promoters are recommended for use when compositions of this invention include siliceous aggregates, such as silica sands, crushed rock and silicates, and alumina based aggregates.
The organofunctional silanes are used in a quantity sufficient to improve adhesion between the resin and aggregate. Typical usage levels of these silanes are 0.1 to 1.5% based on resin weight. Illustrative of silanes that are useful are those represented by the generic Formula (I).
(RO)3-Si-OR
(I) G

The compositions of this invention can include fillers, modifiers, and aggregates which are conventionally used with phenolic resole resins. The aggregate material may be a particulate material such as that in granular, powder, or flake form.
Suitable aggregate materials include but are not limited to: magnesia, magnesite, alumina, zirconia, silica, zircon sand, olivine sand, silicon carbide, silicon nitride, boron nitride, bauxite, quartz, cllromite, and corwdum. For certain applications, low density aggregate materials such as vermiculite, perlite, and pumice are preferred.
For other applications, preferable high density aggregates include: limestone, quartz, sand, gravel, crushed rock, broken brick, and air cooled blast fwnace slag.
Sand, gravel, and crushed rock are preferred aggregates in polymeric concrete.
Fillers such as calcium carbonate, kaolin, mica, wollastonite, and barites can be used in quantities of up to about 50% by weight of the formulated resin product. The quantity of such fillers can equal the quantity of the resin. Hollow microspheres of glass, phenolic resin, or ceramic can also be used in quantities of up to about 20% of the formulated resin prodltCt. Other optional modifiers, particularly in polymer concrete, include fibers such as steel, alkali resistant glass, polyester, carbon, silicon carbide, asbestos, wollastonite fibers, and aromatic polyamides such as KEVLARC~ aramid fiber which is sold by DuPont Advanced Fiber Systems, Richmond , VA, and polypropylene.
The quantity of such fibers can vary over a wide range sufficient to improve the strength of the composition, e.g., from about 2% to 5% by weight of aggregate when aggregate 1S used 111 the Co111pOS1t1011.
The raw batch C0111p051t1On5 produced by combmnlg the hardenable resin binder, aggregate and hardening agent may additionally comprise any of a number of optional modifiers or additives including non-reactive solvents, silanes, hexamethylenetetraamine, clays, graphite, iron oxide, carbon pitch, silicon dioxide, metal powders such as aluminum, magnesium, silicon, surfactants, dispersallts, air detraining agents, and mixtures thereof. Air detraining agents such as antifoamers, e.g., dimethylpolysiloxane and the lilce, can be employed in an amomlt sufficient to increase the strength of the composition. Such quantities can vary over a broad range such as fTOlll abOLlt 0.005% to 0.1% based on the weight of resin and preferably from about 0.01% to 0.05% based on the weight of resin. Illustrative of additional air detraining agents there can be mentioned: various acetylenic derivatives such as the SURFYNOLS of Air Products and Chemicals, Inc., Allentowll, PA, such as SURFYNOL DF-110L, SURFYNOL 104, alld SURFYNOL GA; and various siloxanes such as dimethylpolysiloxane and dimethylsiloxane-allcylene oxide block copolymer such as PS073 which is supplied by United Chemical Technologies, Inc., Bristol, PA.
In foundry applications and sand-binder overlays, or where silica sand is used as the aggregate, a preferred additive is a silane adhesion promoter, such as 3-aminopropyltriethoxysilalle. In refractory applications, clays, metal powders (e.g., aluminum, magnesium, or silicon), and graphite are preferred additives. When graphite or metal powders of alLllllllltl111, lnaglleSlllln, or silicon or mixtures thereof are used as additives, the amolult of aggregate, such as alumina or magnesia, can be reduced to as low as about 70% by weight Of tile C0111pOS1t1O11.
Phenolic Resole Resole resins are thermosetting, i.e., they form an infusible three-dimensional pO1y111eT L117011 appllCatlOll Of heat and are produced by the reaction of a phenol and a molar excess of a phenol-reactive aldehyde typically in the presence of an alkali, alkaline earth, or other metal compound as a condensing catalyst.
The phenolic resole which may be used with tile e111bOd1111e11tS Of the present invention play be obtained by the reaction of a phenol, such as phenol itself, cresol, Te50TC11101, J,5-Xy1e1101, bisphenol-A, other substituted phenols, alld mixtures of ally Of these compounds, with an aldehyde such as, for example, formaldehyde, paraformaldehyde, acetaldehyde, filrfuraldehyde, and mixtures of any of these aldehydes.
A broad range of phenolic resoles in fact may be used with the various e111bOd1111e11tS Of thlS 111Ve11t1O11. These can be phenol-formaldehyde resoles or those where phenol is partially or completely substituted by one or more reactive phenolic compolmds and the aldehyde portion can be partially or wholly replaced by other aldehyde compolulds. The preferred phenolic resole resin is the condensation product of phenol and formaldehyde.
A molar excess of aldehyde per mole of phenol is used to hake the resole resins used in the present inventions. The preferred molar ratio of phenol to aldehyde is in the range of from about 1:2.0 to about 1:2.4. A convenient way to carry out the reaction is by heating the mixture under reflux at atmospheric or reduced pressure conditions. Reflux, however, is not required.
The reaction mixture, is typically heated until from about 80 percent to about 98 percent of the aldehyde has reacted. Although the reaction can be carried out under reflux until about 98 percent of the aldehyde has reacted, prolonged heating is required and it is preferred to continue the heating only until about 80 percent to 90 percent of the aldehyde has reacted. At this point, the reaction mixture is heated under vacuum at a pressure of about 501111n of Hg until the free formaldehyde in the mixture is less than about 1 percent to about 2 percent. Preferably, the reaction is carried out at 95° C until the free formaldehyde is less than about 0.1 percent by weight of the mixtlue. The catalyst may be precipitated from the reaction mixture before the vacuum heating step if desired.
Levels of free phenol and free formaldehyde are measured by methods well 1c11own in the art.
Preparation The preferred phenolic resole used here is a liquid resole having the phenol and formaldehyde ratio of about 1:2.2. The resole is fiuther mixed with silica sand, a magnesilun oxide curing component and water. Lightburned magnesilun oxide particles are preferred at a concentration of between about 0.5% and 50%, and more preferably between about 12% and 18%. The binders were prepared as described below alld tested for compressive strength at room temperature after 3 to 4 hours and then again at 24 flours post preparation. Compressive strength was also tested after the binder samples were heated to 110°C or 125°C in an oven for about 1 hour.
Two standard resole resins were used in the examples that follow. The standard resole resins were Resin 1 having a phenol-to-formaldehyde ratio of 1:2.2, and Resin 2 having a phenol-to-formaldehyde ratio of about 1:0.9. Both resins are conventional liquid resoles. Refractory Mixtures A-D were prepared employing either Resin 1 or Resin ?. The silica sand used in Refiactory Mixtures A-I~
was 60 mesh; however, silica sand of varying grain size may be used. The units for the components of Refractory Mixtures A-D were based on the silica sand content being set to 200 parts and all other components were then set to pal-ts per two hluldred of silica sand. After preparation the damp binders were hand-pressed into 20-gram lltlgget5 Of abOllt 1 alld ~~8-lllCh 111 d1a111eteT alld about J~8-lllch thiclaless. The compressive strengths ofthe nuggets made from Refractory Mixtures A-D are summarized in Table 1.

Refractory Mixture A
Silica sand 200 MAGCHEM 50 1.8 Resin 1 12 Water 4 The silica sand and MAGCHEM 50 were mixed together and then the resin and water were added and mixed and mixed thoroughly for 5 minutes. MAGCHEM
50 is a lightburned magnesium oxide available from Martin Marietta Magnesia Specialties Company, Baltimore, MD. The temperature of the refractory mixture increased from 24°C to 26°C. The material was vibratable in damp form.
Refractory Mixture B
Silica sand 200 MAGCHEM 50 2.4 Resin 1 12 Water 4 The silica sand and MAGCI~EM 50 were mixed then the resin and water were added and mixed thoroughly. The temper attire of the refractory mixture increased from 24°C to 26°C. The material was not vibratable in damp form.
Refiactory Mixture C
Silica Sand 200 MAGCHEM 50 1.5 Resin 1 10 Water 4 The silica sand and MAGCHEM 50 were mixed thoroughly for about 5 minutes. The resin and water were added and mixed thoroughly for 5 minutes.
The temperature of the refractory mixture increased from 24°C to 26°C. The material was vibratable in damp form.
Refractory Mixture D
Silica sand 200 MAGGHEM 50 1.2 Resin 1 8 Water 4 The silica sand and MAGCHEM 50 were mixed together and the resin and water added and mixed thoroughly for 5 minutes. The temperature of the refractory mixture increased from 23°C to 24°C. The material was damp, but did not vibrate well.
Compressive Strength Analysis The Locap Testing Machine manufactured by Tinius Olsen, Willow Grove, PA, was used for the compression strength analysis. The test measures the compression strength parallel to the surface of a specimen ("nugget") by the following steps: (1) the surface of a cylindrical test specimen is carefully ground for smoothness and to provide right angles at the cylindrical edges; (2) the specimen is placed on the machine surface; (3) a vertical load is applied to the flat surface of the specimen at a rate of about 0.1 111Ch per minute; (4) a digital display indicates the load at increasing 111terValS Of S poL111dS Lllltll the specimen fails; and (5) the final load is divided by the sluface area of the specimen to arrive at a compressive strength in units of pounds per square inch ("psi").

Table 1. Compressive Strength of Refractory Mixtures A-C after 3-4, and 24 hours at T00111 temt~erature and after 1 hour at 110°C.
Refractory Mixture / Time l TemperatureSample Compressive Strength, psi A / 4 hours / room temperature 1 347 A / 24 hours / room temperature 1 540 A / 1 hour / 110C 1 1088 B / 3.5 hours / room temperature 1 1254 B / 24 hours / room temperature 1 825 B / 1 hour / 110C after 1 hour at 1 1114 room temperature C / 3 hours l room temperature 1 936 C / 24 hours J loolll temperature 1 921 3 lOG4 C / 1 hour l 110C after 1 hour at 1 1527 room temperatZ.ire D / 3.5 hours / r00111 temperature 1 217 D / 24 hours / roolll telllperattlre 1 504 D / 1 hour / 110C after 1 hour room 1 1877 temperature To test the limitations of the effectiveness of the magnesiwn oxide curing agent in the phenol-to-formaldehyde ratio range of about 1:0.9 to about 1:2.2, the following refractory mixtures, Refractory Mixtures E-I were prepared. In addition to magnesium oxide, a solvent was used. Similar to the preparation of Refractory Mixtures A-D, the units for the components of Refractory Mixtures E-G were based on the silica sand content being set to 100 parts and all other components were than set to parts per 100 of silica sand.
The preparation of Refractory Mixtures E-G was as follows: The silica sand and MAGCHEM 50 were mixed together, then the resin and'y-butyrolactone were added and mixed for 2 minutes. The refractory mixtures were hand-pressed into pills or nuggets of 1-inch diameter and 3/8-111ch thickness.

Refractory Mixture E
Silica sand 100 Resin 2 12 y-butyrolactone 3.21 MAGCHEM 50 1.5 Refractory Mixture r Silica sand 100 MAGCI-IEM 50 1.5 Resin 1 12 'y-butyrolactone 3.22 Refractory Mixture G
Silica sand 100 MAGCHEM 50 1.5 50:50 Resin l :Resia 2 12 'y-butyrolactone 3.24 Refractory Mixture H
Silica sand 200 MAGCHEM 50 1.5 Resin 2 10 Water 4 The units for the components of Refractory Mixture II were based on the silica sand content being set to 200 parts and all other components were then set to parts per two hundred of silica sand. Refractory mixture H was prepared by mixing the silica sand and MAGCHEM 50 thoroughly for 1 minute and then adding the resin and water and mixing thoroughly for 5 minutes. No temperatLUe change occurred. The damp material was vibratable. The mixture was left to stand for 30 minutes prior to being hand pressed into 20-gram nuggets of 1 and 5/8-inch diameter and 3/8-inch thicfaless.
Qualitative results of strength of Refractory Mixtures E-H are slunmarized in Table 2.
Table 2. Strength Results alld Qualitative Observation of Refractory Mixtures E-H
RefractoryPhenol-to- Strength Results and Qualitative Observations Mixture formaldehyde ratio of (at room temperature) the resole F 1:0.9 After 5 hours the mixture was soft and could be broken with fingertips. After 22 hOUrS, the 1111XtLlre WdS f1T111eT, bllt easily broken.

F 1:2.2 After 3 hOLlr S the mixture had hardened.

Compression strength after 3 hours was 242 psi; after 22 flours, 363 psi.

G 1:1.5 After 3 hours the mixture had hardened.

Compression strength after 3 hours was 6.37 psi; after 22 hours, 114 psi.

II 1:0.9 After 3 hours tile 1111XtL1Te WaS SOft. After 7 hours the mixture had hardened slightly, but could be broken with fingertips.
After 24 hOLlrS tile 1111Xt11re defOr111ed alld 110 C0111preSS1011 Strength COLlld be 111eaSL11'ed.

The differences in compressive strength varied SLlrpT1S111g1y and Lulexpectedfy with the phenol-to-formaldehyde ("P/F") ratio in the range of resins tested.
The teachings of the prior art suggested that similar refractory mixtlue strength should be observed with the use of resole resins in the PlF ratio of between about 1:1 and about 1:3. However, the strengths of the refractories produced in these examples are not unrelated to the phenol-to-formaldehyde ratio in the studied range. In fact, the compression strength of refractories in the range of P/F ratios of about 1:1 to about 1:1.5 are sluprisingly lower than the compression strengths of refractories containing resins that have a P/F ratio of between about 1:2.0 and about 1:2.4.
The units of the components of Refractory Mixtures I, J and I~ were based on tile silica sand content being set to 200 parts and all other components are then set to parts per two hundred of silica sand.

Refractory Mixture I
Silica sand 200 Resin 1 16 Water 2 'y-butyrolactone 4.29 The silica sand and MAGCHEM 50 were mixed together, then the resin, water and y-butyrolactone were added and mixed. The mixture was hand-pressed into 20-gram nuggets each of 1 and 5/8-inch diameter with a thiclazess of 3/8-inch.
The material was vibratable and the binder temperature increased from 24°C
to 29°C. The reSLlltS Of the CO111pTe5SlVe Strength allaly5lS Of the blllder are SLL111111ar1Zed in Table 3.
Refractory Mixture ,T
Silica sand 200 MAGCHEM 50 1.5 Resin 1 12 Water 3 'y-butyrolactone 3.22 The silica sand and MAGCIIEM 50 were mixed together, then the resin, water and 'y-butyrolactone were added and mixed. The mixture was hand-pressed into gram nuggets each of 1 alld 5/8-111Ch diameter with a 3/8-inch thiclaless. The material was viUratable. The temperature of the binder increased from 24°C to 29°C. The results of the compressive strength of the binders are summarized in Table 3.
Refractory MixtLUe K
Silica sand 200 Resin 1 8 Water 4 'y-butyrolactone 2.15 The silica sand and MAGCIIEM 50 were mixed thoroughly and then the resin, water and'y-butyrolactone were added and mixed thoroughly for 5 minutes. The temperature of the mixture increased from 24°C to 26°C. The material was vibratable. The mixture was hand-pressed into 20-gram nuggets each of 1 and inch diameter with a thiclazess of 3/8-inch. The results of the compressive strength of the binders are summarized in Table 3.

Table 3. Compressive Strength Results for Refractory Mixtlues I - I~.
Refractory Mixture / Sample Compressive Strength, Time / psi Temperature I / 3 horns / room temperature1 193 I / 24 hours / room temperature1 39G

I / 1 hour, 45 minutes 1 3877 at room temper attire then 1 2 1181 hour, 15 minutes at 125C 3 4060 J / 3 hours / room temperature1 2G3 J / 72 hours / room temperature1 323 J / 1 hour, 45 minutes 1 3136 at r00111 temperature then 1 hour,2 1809 minutes at 125C 3 2424 K / 2 hours / room temperature1 150 K / 3 hours / room temperature1 157 K / 48 hours / room temperature1 270 K / 1 hour, 45 minutes 1 1257 / 125C.

K / 1 hour, 15 minutes 1 1428 examples of further embodiments of the present invention are Refractory Mixtures L-Q. Magnesium oxide aggregate was used in Refractory Mixtures L-Q.
Magnesium oxide aggregates of varying particle sizes are known in the art. The units of the components of Refractory Mixtlues L-Q were based on the magnesium oxide aggregate content being set to 400 parts and all other components were then set to parts per 400 Of 111ag11eS1L1111 OXlde.

Refractory Mixture L
Magnesium oxide 400 Resin 1 40 Water 8 Magnesium oxide (aggregate) and MAGCHEM 50 were mixed for 2 minutes, then the resin and water were added and mixed for 3 minutes uniformly. 45 g samples were hand-pressed into nuggets havnlg a surface area of 2.07 in2 (1 and 5/8-inch diameter). The results of the compressive strength of the binders is summarized in Table 4.
Refractory Mixtlu a M
Magnesium oxide 400 MAGCIIEM 50 2.4 Resin 1 24 Water 16 Magnesium oxide (aggregate) and MAGCHEM 50 were mixed for 2 111111L1teS, tllell tile 1'eSlll alld water were added alld 1111Xed fOT 5 111111L1te5 LllllfOrlnly. The product material was vibratable. 45 g samples were hand-pressed into nuggets. The results of the compressive strength of the binders is summarized in Table 4.
Refractory Mixture N
Magnesium oxide400 MAGCI-iEM 50 1.92 Resin 1 24 Water 16 Magnesium oxide (aggregate) and MAGCHEM 50 were mixed for 5 minutes, then the resin and water were added and mixed for 5 minutes. The mixture was wet initially, then gradually dried out. The product mixture was vibratable. 45 g samples were hand-pressed into nuggets. The results of the compressive strength of the binders is summarized in Table 4.

Refractory Mixture P

Magnesium oxide 400 MAGCHEM 50 1.92 Resin 1 24 Water 10 Magnesium oxide (aggregate) and MAGCHEM 50 were mixed uniformly for 5 minutes, then the resin and water were added and mixed for 5 minutes.
Agglomeration was suppressed resulting in relatively small particle size. The product mixture was very vibratable. 45 g samples were hand-pressed into nuggets. The results of the compressive strength of the binders is swnmarized in Table 4.
Refractory Mixture Q
Magnesium oxide 400 MAGCHEM 50 1.28 Resin 1 16 Water 10 Magnesium oxide (aggregate) and MAGCHEM 50 were mixed thoroughly, then the resin and water were added and mixed for 5 minutes. The mixtL~re had the appearance of damp sand. The product mixture was vibratable. 45 g samples were hand-pressed into nuggets. The results of the compressive strength of the binders is summarized in Table 4.

Table 4. Compressive Strength Results for Refractory Mixtures L - Q
Refiactory Mixture / Sample Compressive Strength, Time / psi ' Temperature L / 3 hours / 100111 1 1229 temperature L / 1 hoLU at room temperature1 12500 then 1 hour at 125C 2 44GG

L / 24 hours / room temperature1 1268 ' M / 3 hours / rOO111 1 744 te111peratllle M / 1 hour at 125C 1 2036 M / 24 hours / r00111 1 928 temperatlue N / 3 hours / room temperature1 485 N / 1 hour at 125C 1 3084 N / 24 hours / room temperature1 534 P / 3 hOLlrS / r00111 1 785 te111peTatllre P / 1 hour at 125C 1 2411 P / 24 hours / room temperature1 1022 Q / 1 hour at 125C 1 182 Q / 24 hours / room temperature1 250 There have been described here various compositions of refractory mixtures CO111p11Slllg reSOle re5111S Collta111111g varying levels of magnesilun oxide as a hardening agent. The prior art teaches that strength of binders using resins having a P/F ratio in the range of 1:1 to 1:3 should result in refractories of similar strength. The results of the experiments described above have 5hOW11 5L1rpr1S111g and Lmexpected results related to the use of resole resins having a P/F ratio in the range of 1:2.0 to about 1:2.4.
With referel~ce to Table 2, the compression strengths ofrefractories prepared using resins of differing P/F ratios are compared. The compression strength of refractories in the range of P/F ratios of about 1:1 to about 1:1.5 are surprisingly lower than the compression strengths of refractories containing resins that have a P/F ratio of between about 1:2.0 and about 1:2.4. No compression strength could effectively be measured for P/F ratios of 1:0.9 (Examples E, II). For P/F ratio 1:1.5 (Example G), IOW C0111pTeS51011 StrellgthS rallglllg fr0111 6 psi after 3 hours at room temperature to 114 psi after 22 hours at room temperature were observed. In contrast, a preferred P/F ratio of 1:2.2 (Example F) provided compression strengths ranging from 242 psi after 3 hours at room temperature to 363 psi after 22 hours at room temperature, an increase of ca. 4000% and 218%, respectively. Using the preferred P/F ratio, unexpected improvements in compression strengths can range from about 200 psi up to >10,000 psi, as described in the foregoing examples. It has been folmd that the resins having a P/F ratio ranging from about 1:2.0 to about 1:2.4 are critical elements of the present invention. A preferred P/F ratio is 1:2.2.
It should be understood that various changes and modifications preferred in the embodiment described herein will be apparent to those skilled in the art.
Such changes and modifications can be made without departing from the spirit and scope of the present invention and without demising the attendant advantages. It is, therefore, intended that such changes and modifications be covered by the appended claims.

Claims

WHAT IS CLAIMED IS:

1. A composition comprising:

a) a liquid resole having a mole ratio of phenol to formaldehyde ranging from about 1:2.0 to about 1: 2.4;
b) an aggregate; and c) a lightburned magnesium hardening agent.

2. The composition of claim 1 which has been hardened at ambient temperature.

3. The composition of claim 1 wherein the liquid resole is present in an amount ranging from about 1 % to about 20% by weight, based on the total weight of the aggregate.

4. The composition of claim 1 wherein the liquid resole is present in an amount ranging from about 4% to about 10% by weight, based on the total weight of the aggregate.

5. The composition of claim 1 wherein the composition further comprises a solvent.

6. The composition of claim 5 wherein the solvent is selected from the group consisting of water, .gamma.-butyrolactone, and mixtures thereof.

7. The composition of claim 5 wherein the solvent is water.

8. The composition of claim 5 wherein the solvent is .gamma.-butyrolactone.

9. The composition of claim 1 wherein the mole ration of phenol to formaldehyde is about 1:2.2.

10. The composition of claim 1 wherein the aggregate is silica sand.

11. The composition of claim 1 wherein the aggregate is magnesium oxide.

12. The composition of claim 1 wherein the lightburned magnesium hardening agent is lightburned magnesium oxide.

13. The composition of claim 3 wherein the lightburned magnesium hardening agent is lightburned magnesium oxide.

14. The composition of claim 13 wherein the lightburned magnesium oxide is present in an amount ranging from about 0.5% to about 50% by weight, based on the total weight of the liquid resole.

15. The composition of claim 13 wherein the lightburned magnesium oxide is present in an amount ranging from about 8% to about 15% by weight, based on the total weight of the liquid resole.

16. The composition of claim 12 wherein the lightburned magnesium oxide has a surface area of about 10 to about 200 square meters per gram.

17. The composition of claim 12 wherein the lightburned magnesium oxide has a surface area of about 50 to about 100 square meters per gram.

18. A binder composition, comprising:
an aggregate;
a liquid resole having a mole ratio of phenol to formaldehyde ranging from about 1:2.0 to about 1:2.4, in an amount ranging from about 1% to about 20% by weight, based on the total weight of the aggregate;
lightburned magnesium oxide in a amount ranging from 0.5%
to about 50% by weight, based on the total weight of the liquid resole.; and a solvent in an amount ranging from 0% to about 25% by weight, based on the total weight of the aggregate.

19. The binder composition of claim 18 wherein the aggregate is silica sand.

20. The binder composition of claim 18 wherein tile aggregate is magnesium oxide.

21. A binder composition, comprising:
magnesium oxide;
a liquid resole having a mole ratio of phenol to formaldehyde of 1:2.2, in an amount ranging from about 4% t0 about 10% by weight, based on the total weight of the aggregate;
lightburned magnesium oxide in an amount ranging from about 8% to about 15% by weight, based on the total weight of the liquid resole.; and water in an amount ranging front 2% to about 8% by weight, based on the total weight of the aggregate.

22. ~A binder composition,comprising:
silica sand;
a liquid resole having a mole ratio of phenol to formaldehyde of 1:2.2, in an amount ranging from about 4% to about 10% by weight, based on the total weight of the aggregate;
lightburned magnesium oxide in an amount ranging from about 12% to about 18% by weight, based on the total weight of the liquid resole.; and water in all amount ranging from 2% to about 8% by weight, based on the total weight of the aggregate.